
from google.colab import files


uploaded = files.upload()

Saving INNHotelsGroup.csv to INNHotelsGroup (1).csv


import numpy as np
import pandas as pd

# import libraries for data visualization
import matplotlib.pyplot as plt
import seaborn as sns


import io

df = pd.read_csv(io.BytesIO(uploaded['INNHotelsGroup.csv']))
print(df)

      Booking_ID  no_of_adults  no_of_children  no_of_weekend_nights  \
0       INN00001             2               0                     1   
1       INN00002             2               0                     2   
2       INN00003             1               0                     2   
3       INN00004             2               0                     0   
4       INN00005             2               0                     1   
...          ...           ...             ...                   ...   
36270   INN36271             3               0                     2   
36271   INN36272             2               0                     1   
36272   INN36273             2               0                     2   
36273   INN36274             2               0                     0   
36274   INN36275             2               0                     1   

       no_of_week_nights type_of_meal_plan  required_car_parking_space  \
0                      2       Meal Plan 1                           0   
1                      3      Not Selected                           0   
2                      1       Meal Plan 1                           0   
3                      2       Meal Plan 1                           0   
4                      1      Not Selected                           0   
...                  ...               ...                         ...   
36270                  6       Meal Plan 1                           0   
36271                  3       Meal Plan 1                           0   
36272                  6       Meal Plan 1                           0   
36273                  3      Not Selected                           0   
36274                  2       Meal Plan 1                           0   

      room_type_reserved  lead_time  arrival_year  arrival_month  \
0            Room_Type 1        224          2017             10   
1            Room_Type 1          5          2018             11   
2            Room_Type 1          1          2018              2   
3            Room_Type 1        211          2018              5   
4            Room_Type 1         48          2018              4   
...                  ...        ...           ...            ...   
36270        Room_Type 4         85          2018              8   
36271        Room_Type 1        228          2018             10   
36272        Room_Type 1        148          2018              7   
36273        Room_Type 1         63          2018              4   
36274        Room_Type 1        207          2018             12   

       arrival_date market_segment_type  repeated_guest  \
0                 2             Offline               0   
1                 6              Online               0   
2                28              Online               0   
3                20              Online               0   
4                11              Online               0   
...             ...                 ...             ...   
36270             3              Online               0   
36271            17              Online               0   
36272             1              Online               0   
36273            21              Online               0   
36274            30             Offline               0   

       no_of_previous_cancellations  no_of_previous_bookings_not_canceled  \
0                                 0                                     0   
1                                 0                                     0   
2                                 0                                     0   
3                                 0                                     0   
4                                 0                                     0   
...                             ...                                   ...   
36270                             0                                     0   
36271                             0                                     0   
36272                             0                                     0   
36273                             0                                     0   
36274                             0                                     0   

       avg_price_per_room  no_of_special_requests booking_status  
0                   65.00                       0   Not_Canceled  
1                  106.68                       1   Not_Canceled  
2                   60.00                       0       Canceled  
3                  100.00                       0       Canceled  
4                   94.50                       0       Canceled  
...                   ...                     ...            ...  
36270              167.80                       1   Not_Canceled  
36271               90.95                       2       Canceled  
36272               98.39                       2   Not_Canceled  
36273               94.50                       0       Canceled  
36274              161.67                       0   Not_Canceled  

[36275 rows x 19 columns]


df.head()


df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 36275 entries, 0 to 36274
Data columns (total 19 columns):
 #   Column                                Non-Null Count  Dtype  
---  ------                                --------------  -----  
 0   Booking_ID                            36275 non-null  object 
 1   no_of_adults                          36275 non-null  int64  
 2   no_of_children                        36275 non-null  int64  
 3   no_of_weekend_nights                  36275 non-null  int64  
 4   no_of_week_nights                     36275 non-null  int64  
 5   type_of_meal_plan                     36275 non-null  object 
 6   required_car_parking_space            36275 non-null  int64  
 7   room_type_reserved                    36275 non-null  object 
 8   lead_time                             36275 non-null  int64  
 9   arrival_year                          36275 non-null  int64  
 10  arrival_month                         36275 non-null  int64  
 11  arrival_date                          36275 non-null  int64  
 12  market_segment_type                   36275 non-null  object 
 13  repeated_guest                        36275 non-null  int64  
 14  no_of_previous_cancellations          36275 non-null  int64  
 15  no_of_previous_bookings_not_canceled  36275 non-null  int64  
 16  avg_price_per_room                    36275 non-null  float64
 17  no_of_special_requests                36275 non-null  int64  
 18  booking_status                        36275 non-null  object 
dtypes: float64(1), int64(13), object(5)
memory usage: 5.3+ MB


df.shape

(36275, 19)


missing_values=pd.isnull(df)
print(missing_values)

       Booking_ID  no_of_adults  no_of_children  no_of_weekend_nights  \
0           False         False           False                 False   
1           False         False           False                 False   
2           False         False           False                 False   
3           False         False           False                 False   
4           False         False           False                 False   
...           ...           ...             ...                   ...   
36270       False         False           False                 False   
36271       False         False           False                 False   
36272       False         False           False                 False   
36273       False         False           False                 False   
36274       False         False           False                 False   

       no_of_week_nights  type_of_meal_plan  required_car_parking_space  \
0                  False              False                       False   
1                  False              False                       False   
2                  False              False                       False   
3                  False              False                       False   
4                  False              False                       False   
...                  ...                ...                         ...   
36270              False              False                       False   
36271              False              False                       False   
36272              False              False                       False   
36273              False              False                       False   
36274              False              False                       False   

       room_type_reserved  lead_time  arrival_year  arrival_month  \
0                   False      False         False          False   
1                   False      False         False          False   
2                   False      False         False          False   
3                   False      False         False          False   
4                   False      False         False          False   
...                   ...        ...           ...            ...   
36270               False      False         False          False   
36271               False      False         False          False   
36272               False      False         False          False   
36273               False      False         False          False   
36274               False      False         False          False   

       arrival_date  market_segment_type  repeated_guest  \
0             False                False           False   
1             False                False           False   
2             False                False           False   
3             False                False           False   
4             False                False           False   
...             ...                  ...             ...   
36270         False                False           False   
36271         False                False           False   
36272         False                False           False   
36273         False                False           False   
36274         False                False           False   

       no_of_previous_cancellations  no_of_previous_bookings_not_canceled  \
0                             False                                 False   
1                             False                                 False   
2                             False                                 False   
3                             False                                 False   
4                             False                                 False   
...                             ...                                   ...   
36270                         False                                 False   
36271                         False                                 False   
36272                         False                                 False   
36273                         False                                 False   
36274                         False                                 False   

       avg_price_per_room  no_of_special_requests  booking_status  
0                   False                   False           False  
1                   False                   False           False  
2                   False                   False           False  
3                   False                   False           False  
4                   False                   False           False  
...                   ...                     ...             ...  
36270               False                   False           False  
36271               False                   False           False  
36272               False                   False           False  
36273               False                   False           False  
36274               False                   False           False  

[36275 rows x 19 columns]


df.isnull().sum()

Booking_ID                              0
no_of_adults                            0
no_of_children                          0
no_of_weekend_nights                    0
no_of_week_nights                       0
type_of_meal_plan                       0
required_car_parking_space              0
room_type_reserved                      0
lead_time                               0
arrival_year                            0
arrival_month                           0
arrival_date                            0
market_segment_type                     0
repeated_guest                          0
no_of_previous_cancellations            0
no_of_previous_bookings_not_canceled    0
avg_price_per_room                      0
no_of_special_requests                  0
booking_status                          0
dtype: int64


summary = df.describe()
print(summary)

       no_of_adults  no_of_children  no_of_weekend_nights  no_of_week_nights  \
count  36275.000000    36275.000000          36275.000000       36275.000000   
mean       1.844962        0.105279              0.810724           2.204300   
std        0.518715        0.402648              0.870644           1.410905   
min        0.000000        0.000000              0.000000           0.000000   
25%        2.000000        0.000000              0.000000           1.000000   
50%        2.000000        0.000000              1.000000           2.000000   
75%        2.000000        0.000000              2.000000           3.000000   
max        4.000000       10.000000              7.000000          17.000000   

       required_car_parking_space     lead_time  arrival_year  arrival_month  \
count                36275.000000  36275.000000  36275.000000   36275.000000   
mean                     0.030986     85.232557   2017.820427       7.423653   
std                      0.173281     85.930817      0.383836       3.069894   
min                      0.000000      0.000000   2017.000000       1.000000   
25%                      0.000000     17.000000   2018.000000       5.000000   
50%                      0.000000     57.000000   2018.000000       8.000000   
75%                      0.000000    126.000000   2018.000000      10.000000   
max                      1.000000    443.000000   2018.000000      12.000000   

       arrival_date  repeated_guest  no_of_previous_cancellations  \
count  36275.000000    36275.000000                  36275.000000   
mean      15.596995        0.025637                      0.023349   
std        8.740447        0.158053                      0.368331   
min        1.000000        0.000000                      0.000000   
25%        8.000000        0.000000                      0.000000   
50%       16.000000        0.000000                      0.000000   
75%       23.000000        0.000000                      0.000000   
max       31.000000        1.000000                     13.000000   

       no_of_previous_bookings_not_canceled  avg_price_per_room  \
count                          36275.000000        36275.000000   
mean                               0.153411          103.423539   
std                                1.754171           35.089424   
min                                0.000000            0.000000   
25%                                0.000000           80.300000   
50%                                0.000000           99.450000   
75%                                0.000000          120.000000   
max                               58.000000          540.000000   

       no_of_special_requests  
count            36275.000000  
mean                 0.619655  
std                  0.786236  
min                  0.000000  
25%                  0.000000  
50%                  0.000000  
75%                  1.000000  
max                  5.000000


df.describe(include = ['object']).T


# Check for duplicates
duplicate_rows = df[df.duplicated()]

# Print the duplicate rows
print("Duplicate Rows:")
print(duplicate_rows)

# Optionally, you can also count the number of duplicates
num_duplicates = df.duplicated().sum()
print("Number of duplicates:", num_duplicates)

Duplicate Rows:
Empty DataFrame
Columns: [Booking_ID, no_of_adults, no_of_children, no_of_weekend_nights, no_of_week_nights, type_of_meal_plan, required_car_parking_space, room_type_reserved, lead_time, arrival_year, arrival_month, arrival_date, market_segment_type, repeated_guest, no_of_previous_cancellations, no_of_previous_bookings_not_canceled, avg_price_per_room, no_of_special_requests, booking_status]
Index: []
Number of duplicates: 0


import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# Load the dataset
df = pd.DataFrame(df)

# Display basic statistics for numeric variables
numeric_summary = df.describe()
print(numeric_summary)

# Display basic statistics for categorical variables
categorical_summary = df.describe(include=['object'])
print(categorical_summary)

# Define numeric columns to plot
numeric_cols = ['no_of_adults', 'no_of_children', 'no_of_weekend_nights', 'no_of_week_nights', 'lead_time',
                'arrival_year', 'arrival_month', 'arrival_date', 'no_of_previous_cancellations',
                'no_of_previous_bookings_not_canceled', 'avg_price_per_room', 'no_of_special_requests']

# Histograms for numeric columns
plt.figure(figsize=(20, 15))
for i, col in enumerate(numeric_cols, 1):
    plt.subplot(4, 3, i)
    sns.histplot(df[col], kde=True)
    plt.title(f'Histogram of {col}')
    plt.tight_layout()

plt.show()

# Box plots for numeric columns
plt.figure(figsize=(20, 15))
for i, col in enumerate(numeric_cols, 1):
    plt.subplot(4, 3, i)
    sns.boxplot(y=df[col])
    plt.title(f'Box plot of {col}')
    plt.tight_layout()

plt.show()

# Define categorical columns to plot
categorical_cols = ['type_of_meal_plan', 'room_type_reserved', 'market_segment_type', 'booking_status']

# Bar plots for categorical variables
plt.figure(figsize=(15, 10))
for i, col in enumerate(categorical_cols, 1):
    plt.subplot(2, 2, i)
    df[col].value_counts().plot(kind='bar')
    plt.title(f'Bar plot of {col}')
    plt.tight_layout()

plt.show()

       no_of_adults  no_of_children  no_of_weekend_nights  no_of_week_nights  \
count  36275.000000    36275.000000          36275.000000       36275.000000   
mean       1.844962        0.105279              0.810724           2.204300   
std        0.518715        0.402648              0.870644           1.410905   
min        0.000000        0.000000              0.000000           0.000000   
25%        2.000000        0.000000              0.000000           1.000000   
50%        2.000000        0.000000              1.000000           2.000000   
75%        2.000000        0.000000              2.000000           3.000000   
max        4.000000       10.000000              7.000000          17.000000   

       required_car_parking_space     lead_time  arrival_year  arrival_month  \
count                36275.000000  36275.000000  36275.000000   36275.000000   
mean                     0.030986     85.232557   2017.820427       7.423653   
std                      0.173281     85.930817      0.383836       3.069894   
min                      0.000000      0.000000   2017.000000       1.000000   
25%                      0.000000     17.000000   2018.000000       5.000000   
50%                      0.000000     57.000000   2018.000000       8.000000   
75%                      0.000000    126.000000   2018.000000      10.000000   
max                      1.000000    443.000000   2018.000000      12.000000   

       arrival_date  repeated_guest  no_of_previous_cancellations  \
count  36275.000000    36275.000000                  36275.000000   
mean      15.596995        0.025637                      0.023349   
std        8.740447        0.158053                      0.368331   
min        1.000000        0.000000                      0.000000   
25%        8.000000        0.000000                      0.000000   
50%       16.000000        0.000000                      0.000000   
75%       23.000000        0.000000                      0.000000   
max       31.000000        1.000000                     13.000000   

       no_of_previous_bookings_not_canceled  avg_price_per_room  \
count                          36275.000000        36275.000000   
mean                               0.153411          103.423539   
std                                1.754171           35.089424   
min                                0.000000            0.000000   
25%                                0.000000           80.300000   
50%                                0.000000           99.450000   
75%                                0.000000          120.000000   
max                               58.000000          540.000000   

       no_of_special_requests  
count            36275.000000  
mean                 0.619655  
std                  0.786236  
min                  0.000000  
25%                  0.000000  
50%                  0.000000  
75%                  1.000000  
max                  5.000000  
       Booking_ID type_of_meal_plan room_type_reserved market_segment_type  \
count       36275             36275              36275               36275   
unique      36275                 4                  7                   5   
top      INN00001       Meal Plan 1        Room_Type 1              Online   
freq            1             27835              28130               23214   

       booking_status  
count           36275  
unique              2  
top      Not_Canceled  
freq            24390


import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# Load the dataset
df = pd.DataFrame(df)

# Define numeric and categorical columns
numeric_cols = ['no_of_adults', 'no_of_children', 'no_of_weekend_nights', 'no_of_week_nights', 'lead_time',
                'arrival_year', 'arrival_month', 'arrival_date', 'no_of_previous_cancellations',
                'no_of_previous_bookings_not_canceled', 'avg_price_per_room', 'no_of_special_requests']
categorical_cols = ['type_of_meal_plan', 'room_type_reserved', 'market_segment_type', 'booking_status']

# Plotting correlation matrix for numeric variables
plt.figure(figsize=(12, 8))
correlation_matrix = df[numeric_cols].corr()
sns.heatmap(correlation_matrix, annot=True, cmap='coolwarm', vmin=-1, vmax=1)
plt.title('Correlation Matrix of Numeric Variables')
plt.show()

# Pairplots for numeric variables
sns.pairplot(df[numeric_cols])
plt.suptitle('Pairplots of Numeric Variables', y=1.02)
plt.show()

# Box plots of numeric variables against categorical variables
for cat_col in categorical_cols:
    for num_col in numeric_cols:
        plt.figure(figsize=(10, 6))
        sns.boxplot(x=df[cat_col], y=df[num_col])
        plt.title(f'Box plot of {num_col} by {cat_col}')
        plt.xticks(rotation=90)
        plt.show()

# Bar plots of categorical variables against each other
for i, cat_col1 in enumerate(categorical_cols):
    for j, cat_col2 in enumerate(categorical_cols):
        if i < j:
            plt.figure(figsize=(10, 6))
            sns.countplot(x=df[cat_col1], hue=df[cat_col2])
            plt.title(f'Count plot of {cat_col1} by {cat_col2}')
            plt.xticks(rotation=90)
            plt.show()


import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# Load the dataset
df=pd.DataFrame(df)

# Calculate the number of bookings for each month
monthly_bookings = df['arrival_month'].value_counts().sort_index()

# Prepare the data for plotting
monthly_bookings_df = monthly_bookings.reset_index()
monthly_bookings_df.columns = ['Month', 'Number of Bookings']

# Plot the number of bookings for each month
plt.figure(figsize=(10, 6))
sns.barplot(data=monthly_bookings_df, x='Month', y='Number of Bookings', palette='viridis', hue='Month', dodge=False)
plt.title('Number of Bookings per Month')
plt.xlabel('Month')
plt.ylabel('Number of Bookings')
plt.xticks(monthly_bookings_df['Month'] - 1, ['Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec'])
plt.legend([],[], frameon=False)
plt.show()

# Print the monthly bookings
print(monthly_bookings)

arrival_month
1     1014
2     1704
3     2358
4     2736
5     2598
6     3203
7     2920
8     3813
9     4611
10    5317
11    2980
12    3021
Name: count, dtype: int64


import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# Load the dataset
df=pd.DataFrame(df)


# Calculate the number of bookings for each market segment
market_segment_bookings = df['market_segment_type'].value_counts()

# Prepare the data for plotting
market_segment_df = market_segment_bookings.reset_index()
market_segment_df.columns = ['Market Segment', 'Number of Bookings']

# Plot the number of bookings for each market segment
plt.figure(figsize=(10, 6))
sns.barplot(data=market_segment_df, x='Market Segment', y='Number of Bookings', palette='viridis', hue='Market Segment', dodge=False)
plt.title('Number of Bookings by Market Segment')
plt.xlabel('Market Segment')
plt.ylabel('Number of Bookings')
plt.xticks(rotation=45)
plt.legend([],[], frameon=False)
plt.show()

# Print the market segment bookings
print(market_segment_bookings)

market_segment_type
Online           23214
Offline          10528
Corporate         2017
Complementary      391
Aviation           125
Name: count, dtype: int64


import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# Load the dataset
# Load the dataset
df=pd.DataFrame(df)


# Calculate the average room price for each market segment
average_price_per_segment = df.groupby('market_segment_type')['avg_price_per_room'].mean().reset_index()

# Plot the average room price for each market segment
plt.figure(figsize=(10, 6))
sns.barplot(data=average_price_per_segment, x='market_segment_type', y='avg_price_per_room', palette='viridis', hue='market_segment_type', dodge=False)
plt.title('Average Room Price by Market Segment')
plt.xlabel('Market Segment')
plt.ylabel('Average Room Price')
plt.xticks(rotation=45)
plt.legend([],[], frameon=False)
plt.show()

# Print the average room prices
print(average_price_per_segment)

  market_segment_type  avg_price_per_room
0            Aviation          100.704000
1       Complementary            3.141765
2           Corporate           82.911740
3             Offline           91.632679
4              Online          112.256855


import pandas as pd

# Load the dataset
df=pd.DataFrame(df)

# Calculate the total number of bookings
total_bookings = df.shape[0]

# Calculate the number of canceled bookings
canceled_bookings = df[df['booking_status'] == 'Canceled'].shape[0]

# Calculate the percentage of canceled bookings
percentage_canceled = (canceled_bookings / total_bookings) * 100

# Print the result
print(f'Percentage of canceled bookings: {percentage_canceled:.2f}%')

Percentage of canceled bookings: 32.76%


import pandas as pd

# Load the dataset
df=pd.DataFrame(df)

# Calculate the total number of repeating guests
total_repeating_guests = df[df['repeated_guest'] == 1].shape[0]

# Calculate the number of canceled bookings among repeating guests
canceled_repeating_guests = df[(df['repeated_guest'] == 1) & (df['booking_status'] == 'Canceled')].shape[0]

# Calculate the percentage of canceled bookings among repeating guests
percentage_canceled_repeating_guests = (canceled_repeating_guests / total_repeating_guests) * 100

# Print the result
print(f'Percentage of canceled bookings among repeating guests: {percentage_canceled_repeating_guests:.2f}%')

Percentage of canceled bookings among repeating guests: 1.72%


import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
from scipy.stats import chi2_contingency

# Load the dataset
df=pd.DataFrame(df)

# Calculate the cancellation rate for each number of special requests
special_requests_cancellation = df.groupby('no_of_special_requests')['booking_status'].value_counts(normalize=True).unstack().fillna(0)
special_requests_cancellation['Cancellation_Rate'] = special_requests_cancellation['Canceled'] * 100

# Prepare the data for plotting
special_requests_cancellation_df = special_requests_cancellation.reset_index()

# Plot the cancellation rate by number of special requests
plt.figure(figsize=(10, 6))
sns.barplot(data=special_requests_cancellation_df, x='no_of_special_requests', y='Cancellation_Rate', hue='no_of_special_requests', dodge=False, palette='viridis')
plt.title('Cancellation Rate by Number of Special Requests')
plt.xlabel('Number of Special Requests')
plt.ylabel('Cancellation Rate (%)')
plt.legend([],[], frameon=False)
plt.show()

# Perform a Chi-Square test for independence
contingency_table = pd.crosstab(df['no_of_special_requests'], df['booking_status'])
chi2, p, dof, ex = chi2_contingency(contingency_table)

# Print the result of the Chi-Square test
print(f'Chi-Square Test: chi2={chi2}, p-value={p}')

# Interpretation based on p-value
if p < 0.05:
    print('The number of special requests has a significant effect on booking cancellation (p < 0.05).')
else:
    print('The number of special requests does not have a significant effect on booking cancellation (p >= 0.05).')

Chi-Square Test: chi2=2421.6187208019905, p-value=0.0
The number of special requests has a significant effect on booking cancellation (p < 0.05).


df.isnull().sum()

Booking_ID                              0
no_of_adults                            0
no_of_children                          0
no_of_weekend_nights                    0
no_of_week_nights                       0
type_of_meal_plan                       0
required_car_parking_space              0
room_type_reserved                      0
lead_time                               0
arrival_year                            0
arrival_month                           0
arrival_date                            0
market_segment_type                     0
repeated_guest                          0
no_of_previous_cancellations            0
no_of_previous_bookings_not_canceled    0
avg_price_per_room                      0
no_of_special_requests                  0
booking_status                          0
dtype: int64


import pandas as pd
from sklearn.preprocessing import OneHotEncoder, MinMaxScaler

# Load the dataset
df=pd.DataFrame(df)

# Create a copy of the dataset
df1 = df.copy()

# Step 3: Encoding Categorical Variables
categorical_cols = ['type_of_meal_plan', 'room_type_reserved', 'market_segment_type']
encoder = OneHotEncoder(sparse_output=False, drop='first')  # Use sparse_output instead of sparse
encoded_features = encoder.fit_transform(df1[categorical_cols])
encoded_feature_names = encoder.get_feature_names_out(categorical_cols)

# Convert encoded features to a DataFrame
encoded_df = pd.DataFrame(encoded_features, columns=encoded_feature_names)

# Concatenate encoded features with the original dataset
data_encoded = pd.concat([df1.drop(columns=categorical_cols), encoded_df], axis=1)

# Step 4: Creating Interaction Features
data_encoded['total_nights'] = data_encoded['no_of_weekend_nights'] + data_encoded['no_of_week_nights']

# Step 5: Scaling and Normalization
scaler = MinMaxScaler()
scaled_features = scaler.fit_transform(data_encoded)

# Convert scaled features to a DataFrame
data_scaled = pd.DataFrame(scaled_features, columns=data_encoded.columns)

# Display the processed DataFrame
data_scaled.head()


import pandas as pd
from sklearn.preprocessing import OneHotEncoder, MinMaxScaler

# Load the dataset
df=pd.DataFrame(df)

# Create a copy of the dataset
df1 = df.copy()

# Step 1: Create New Features
# Example: Total nights stayed
df1['total_nights'] = df1['no_of_weekend_nights'] + df1['no_of_week_nights']

# Example: Average price per night
df1['total_nights'] = df1['total_nights'].replace(0, 1)  # Avoid division by zero by replacing 0 nights with 1
df1['avg_price_per_night'] = df1['avg_price_per_room'] / df1['total_nights']

# Drop the original columns used to create new features if necessary
# df1.drop(columns=['no_of_weekend_nights', 'no_of_week_nights'], inplace=True)

# Step 2: Encode Categorical Variables
categorical_cols = ['type_of_meal_plan', 'room_type_reserved', 'market_segment_type']
encoder = OneHotEncoder(sparse_output=False, drop='first')
encoded_features = encoder.fit_transform(df1[categorical_cols])
encoded_feature_names = encoder.get_feature_names_out(categorical_cols)

# Convert encoded features to a DataFrame
encoded_df = pd.DataFrame(encoded_features, columns=encoded_feature_names)

# Concatenate encoded features with the original dataset
data_encoded = pd.concat([df1.drop(columns=categorical_cols), encoded_df], axis=1)

# Step 3: Scaling and Normalization
scaler = MinMaxScaler()
scaled_features = scaler.fit_transform(data_encoded)

# Convert scaled features to a DataFrame
data_scaled = pd.DataFrame(scaled_features, columns=data_encoded.columns)

# Display the processed DataFrame
data_scaled.head()


import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from sklearn.preprocessing import OneHotEncoder, MinMaxScaler

# Load the dataset
df=pd.DataFrame(df)

# Create a copy of the dataset
df1 = df.copy()

# Step 1: Outlier Detection using Boxplot
numeric_columns = df1.select_dtypes(include=np.number).columns.tolist()
num_columns = len(numeric_columns)

# Determine the grid size
grid_size = int(np.ceil(np.sqrt(num_columns)))

plt.figure(figsize=(15, 12))
for i, variable in enumerate(numeric_columns):
    plt.subplot(grid_size, grid_size, i + 1)
    plt.boxplot(df1[variable], whis=1.5)
    plt.tight_layout()
    plt.title(variable)
plt.show()

# Step 2: Treating Outliers
# Define a function to cap outliers
def cap_outliers(df, column, upper_quantile=0.95):
    upper_limit = df[column].quantile(upper_quantile)
    df[column] = np.where(df[column] > upper_limit, upper_limit, df[column])
    return df

# Apply the function to relevant numeric columns
columns_to_cap = ['lead_time', 'no_of_previous_cancellations', 'no_of_previous_bookings_not_canceled', 'avg_price_per_room', 'no_of_special_requests']

for column in columns_to_cap:
    df1 = cap_outliers(df1, column)

# Step 3: Create New Features
# Example: Total nights stayed
df1['total_nights'] = df1['no_of_weekend_nights'] + df1['no_of_week_nights']

# Example: Average price per night
df1['total_nights'] = df1['total_nights'].replace(0, 1)  # Avoid division by zero by replacing 0 nights with 1
df1['avg_price_per_night'] = df1['avg_price_per_room'] / df1['total_nights']

# Step 4: Encode Categorical Variables
categorical_cols = ['type_of_meal_plan', 'room_type_reserved', 'market_segment_type']
encoder = OneHotEncoder(sparse_output=False, drop='first')
encoded_features = encoder.fit_transform(df1[categorical_cols])
encoded_feature_names = encoder.get_feature_names_out(categorical_cols)

# Convert encoded features to a DataFrame
encoded_df = pd.DataFrame(encoded_features, columns=encoded_feature_names)

# Concatenate encoded features with the original dataset
data_encoded = pd.concat([df1.drop(columns=categorical_cols), encoded_df], axis=1)

# Step 5: Scaling and Normalization
scaler = MinMaxScaler()
scaled_features = scaler.fit_transform(data_encoded)

# Convert scaled features to a DataFrame
data_scaled = pd.DataFrame(scaled_features, columns=data_encoded.columns)

# Display the processed DataFrame
data_scaled.head()


import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.preprocessing import StandardScaler

# Load the manipulated dataset
df=pd.DataFrame(df)
df1 = df.copy()

# Fill missing values for numerical columns with the mean
numerical_cols = df1.select_dtypes(include=['float64', 'int64']).columns
df1[numerical_cols] = df1[numerical_cols].fillna(df1[numerical_cols].mean())

# Fill missing values for categorical columns with the mode
categorical_cols = df1.select_dtypes(include=['object']).columns
df1[categorical_cols] = df1[categorical_cols].fillna(df1[categorical_cols].mode().iloc[0])

# Create dummy variables for categorical columns
df1 = pd.get_dummies(df1, columns=categorical_cols, drop_first=True)

# Scale the features
scaler = StandardScaler()
df1[numerical_cols] = scaler.fit_transform(df1[numerical_cols])

# Step 1: Summary Statistics
numerical_summary = df1.describe()
print("Numerical Summary:\n", numerical_summary)

# Step 2: Distribution Plots
plt.figure(figsize=(20, 15))
for i, col in enumerate(numerical_cols, 1):
    plt.subplot(5, 4, i)
    sns.histplot(df1[col], kde=True)
    plt.title(f'Distribution of {col}')
plt.tight_layout()
plt.show()

# Step 3: Correlation Matrix
plt.figure(figsize=(15, 10))
correlation_matrix = df1[numerical_cols].corr()
sns.heatmap(correlation_matrix, annot=True, cmap='coolwarm', fmt='.2f')
plt.title('Correlation Matrix')
plt.show()

# Step 4: Box Plots
plt.figure(figsize=(20, 15))
for i, col in enumerate(numerical_cols, 1):
    plt.subplot(5, 4, i)
    sns.boxplot(x=df1[col])
    plt.title(f'Box Plot of {col}')
plt.tight_layout()
plt.show()

# Step 5: Count Plots
categorical_cols = df1.select_dtypes(include=['uint8']).columns
plt.figure(figsize=(20, 15))
for i, col in enumerate(categorical_cols, 1):
    plt.subplot(5, 4, i)
    sns.countplot(y=df1[col])
    plt.title(f'Count Plot of {col}')
plt.tight_layout()
plt.show()

Numerical Summary:
        no_of_adults  no_of_children  no_of_weekend_nights  no_of_week_nights  \
count  3.627500e+04    3.627500e+04          3.627500e+04       3.627500e+04   
mean   4.270112e-17    1.518044e-17          9.950536e-17      -1.165466e-16   
std    1.000014e+00    1.000014e+00          1.000014e+00       1.000014e+00   
min   -3.556844e+00   -2.614704e-01         -9.311902e-01      -1.562353e+00   
25%    2.988926e-01   -2.614704e-01         -9.311902e-01      -8.535778e-01   
50%    2.988926e-01   -2.614704e-01          2.174012e-01      -1.448030e-01   
75%    2.988926e-01   -2.614704e-01          1.365993e+00       5.639718e-01   
max    4.154629e+00    2.457446e+01          7.108950e+00       1.048682e+01   

       required_car_parking_space     lead_time  arrival_year  arrival_month  \
count                3.627500e+04  3.627500e+04  3.627500e+04   3.627500e+04   
mean                 3.917534e-17  6.463931e-17 -2.254506e-13   1.436266e-16   
std                  1.000014e+00  1.000014e+00  1.000014e+00   1.000014e+00   
min                 -1.788193e-01 -9.918878e-01 -2.137469e+00  -2.092496e+00   
25%                 -1.788193e-01 -7.940515e-01  4.678430e-01  -7.895014e-01   
50%                 -1.788193e-01 -3.285544e-01  4.678430e-01   1.877443e-01   
75%                 -1.788193e-01  4.744282e-01  4.678430e-01   8.392415e-01   
max                  5.592239e+00  4.163493e+00  4.678430e-01   1.490739e+00   

       arrival_date  repeated_guest  no_of_previous_cancellations  \
count  3.627500e+04    3.627500e+04                  3.627500e+04   
mean  -4.701041e-17    1.704127e-17                  1.008765e-17   
std    1.000014e+00    1.000014e+00                  1.000014e+00   
min   -1.670074e+00   -1.622099e-01                 -6.339327e-02   
25%   -8.691889e-01   -1.622099e-01                 -6.339327e-02   
50%    4.610867e-02   -1.622099e-01                 -6.339327e-02   
75%    8.469940e-01   -1.622099e-01                 -6.339327e-02   
max    1.762292e+00    6.164850e+00                  3.523139e+01   

       no_of_previous_bookings_not_canceled  avg_price_per_room  \
count                          3.627500e+04        3.627500e+04   
mean                          -3.094852e-17       -7.051561e-17   
std                            1.000014e+00        1.000014e+00   
min                           -8.745646e-02       -2.947468e+00   
25%                           -8.745646e-02       -6.589979e-01   
50%                           -8.745646e-02       -1.132419e-01   
75%                           -8.745646e-02        4.724127e-01   
max                            3.297706e+01        1.244200e+01   

       no_of_special_requests  
count            3.627500e+04  
mean             1.664952e-17  
std              1.000014e+00  
min             -7.881400e-01  
25%             -7.881400e-01  
50%             -7.881400e-01  
75%              4.837605e-01  
max              5.571362e+00

<Figure size 2000x1500 with 0 Axes>


import pandas as pd
import numpy as np
from statsmodels.stats.outliers_influence import variance_inflation_factor

# Assuming df1 is your preprocessed DataFrame with dummy variables
# Load the manipulated dataset
df=pd.DataFrame(df)
df1 = df.copy()


# Select numeric features for VIF calculation
numeric_cols = df1.select_dtypes(include=['float64', 'int64']).columns

# Calculate VIF for each numeric feature
vif_data = pd.DataFrame()
vif_data['Feature'] = numeric_cols
vif_data['VIF'] = [variance_inflation_factor(df1[numeric_cols].values, i) for i in range(len(numeric_cols))]

print("Variance Inflation Factors (VIF):\n", vif_data)

Variance Inflation Factors (VIF):
                                  Feature        VIF
0                           no_of_adults  16.448306
1                         no_of_children   1.242407
2                   no_of_weekend_nights   1.959993
3                      no_of_week_nights   3.678980
4             required_car_parking_space   1.062769
5                              lead_time   2.174975
6                           arrival_year  29.448847
7                          arrival_month   7.158118
8                           arrival_date   4.204407
9                         repeated_guest   1.595823
10          no_of_previous_cancellations   1.337686
11  no_of_previous_bookings_not_canceled   1.603672
12                    avg_price_per_room  12.751692
13                no_of_special_requests   1.797588


import pandas as pd
import statsmodels.api as sm
from statsmodels.stats.outliers_influence import variance_inflation_factor
from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_absolute_error, mean_squared_error, r2_score
import numpy as np

# Function to calculate VIF
def checking_vif(predictors):
    vif = pd.DataFrame()
    vif["feature"] = predictors.columns
    vif["VIF"] = [round(variance_inflation_factor(predictors.values, i), 2) for i in range(predictors.shape[1])]
    return vif

# Function to compute adjusted R-squared
def adj_r2_score(predictors, targets, predictions):
    r2 = r2_score(targets, predictions)
    n = predictors.shape[0]
    k = predictors.shape[1]
    return 1 - ((1 - r2) * (n - 1) / (n - k - 1))

# Function to compute MAPE
def mape_score(targets, predictions):
    return np.mean(np.abs(targets - predictions) / targets) * 100

# Function to compute different metrics to check performance of a regression model
def model_performance_regression(model, predictors, target):
    pred = model.predict(predictors)
    r2 = r2_score(target, pred)
    adjr2 = adj_r2_score(predictors, target, pred)
    rmse = np.sqrt(mean_squared_error(target, pred))
    mae = mean_absolute_error(target, pred)
    mape = mape_score(target, pred)
    df_perf = pd.DataFrame(
        {
            "RMSE": [rmse],
            "MAE": [mae],
            "R-squared": [r2],
            "Adj. R-squared": [adjr2],
            "MAPE": [mape],
        }
    )
    return df_perf

# Function to iteratively remove multicollinearity
def remove_multicollinearity(data, target, features, threshold=5):
    X = data[features].fillna(data[features].mean())
    X = sm.add_constant(X)
    y = data[target]

    dropped_features = []

    while True:
        vif_df = checking_vif(X)
        print("VIF values:\n", vif_df)

        max_vif = vif_df['VIF'].max()
        if max_vif <= threshold:
            break

        feature_to_drop = vif_df.sort_values('VIF', ascending=False).iloc[0]['feature']
        if feature_to_drop == 'const':
            print("High VIF due to intercept, stopping removal.")
            break

        print(f"Dropping feature with highest VIF: {feature_to_drop}")
        X = X.drop(columns=[feature_to_drop])
        dropped_features.append(feature_to_drop)

    # Split data into training and testing sets
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

    # Train final OLS model
    final_model = sm.OLS(y_train, X_train).fit()

    # Evaluate model performance on training set
    train_perf = model_performance_regression(final_model, X_train, y_train)
    print("Training Performance\n", train_perf)

    # Evaluate model performance on test set
    test_perf = model_performance_regression(final_model, X_test, y_test)
    print("Test Performance\n", test_perf)

    print("Final VIF values:\n", checking_vif(X))
    return X, dropped_features, final_model

# Load and prepare the dataset
df=pd.DataFrame(df)
df1 = df.copy()

# Define target and features
target = 'avg_price_per_room'  # Assuming we want to predict the average price per room
features = ['no_of_adults', 'no_of_children', 'no_of_weekend_nights', 'no_of_week_nights', 'required_car_parking_space',
            'lead_time', 'arrival_year', 'arrival_month', 'arrival_date', 'repeated_guest',
            'no_of_previous_cancellations', 'no_of_previous_bookings_not_canceled', 'no_of_special_requests']

# Remove multicollinearity
final_X, dropped_features, final_model = remove_multicollinearity(df1, target, features)

print("Dropped features to remove multicollinearity:", dropped_features)
print("Final model summary:\n", final_model.summary())

VIF values:
                                  feature          VIF
0                                  const  33396745.53
1                           no_of_adults         1.11
2                         no_of_children         1.03
3                   no_of_weekend_nights         1.05
4                      no_of_week_nights         1.07
5             required_car_parking_space         1.03
6                              lead_time         1.14
7                           arrival_year         1.21
8                          arrival_month         1.22
9                           arrival_date         1.00
10                        repeated_guest         1.54
11          no_of_previous_cancellations         1.33
12  no_of_previous_bookings_not_canceled         1.59
13                no_of_special_requests         1.11
High VIF due to intercept, stopping removal.
Training Performance
         RMSE       MAE  R-squared  Adj. R-squared  MAPE
0  29.860042  22.50379   0.277818        0.277469   inf
Test Performance
         RMSE        MAE  R-squared  Adj. R-squared  MAPE
0  29.965731  22.452745   0.262503        0.261077   inf
Final VIF values:
                                  feature          VIF
0                                  const  33396745.53
1                           no_of_adults         1.11
2                         no_of_children         1.03
3                   no_of_weekend_nights         1.05
4                      no_of_week_nights         1.07
5             required_car_parking_space         1.03
6                              lead_time         1.14
7                           arrival_year         1.21
8                          arrival_month         1.22
9                           arrival_date         1.00
10                        repeated_guest         1.54
11          no_of_previous_cancellations         1.33
12  no_of_previous_bookings_not_canceled         1.59
13                no_of_special_requests         1.11
Dropped features to remove multicollinearity: []
Final model summary:
                             OLS Regression Results                            
==============================================================================
Dep. Variable:     avg_price_per_room   R-squared:                       0.278
Model:                            OLS   Adj. R-squared:                  0.277
Method:                 Least Squares   F-statistic:                     858.3
Date:                Thu, 11 Jul 2024   Prob (F-statistic):               0.00
Time:                        22:31:18   Log-Likelihood:            -1.3974e+05
No. Observations:               29020   AIC:                         2.795e+05
Df Residuals:                   29006   BIC:                         2.796e+05
Df Model:                          13                                         
Covariance Type:            nonrobust                                         
========================================================================================================
                                           coef    std err          t      P>|t|      [0.025      0.975]
--------------------------------------------------------------------------------------------------------
const                                -3.714e+04   1013.185    -36.660      0.000   -3.91e+04   -3.52e+04
no_of_adults                            18.2885      0.356     51.392      0.000      17.591      18.986
no_of_children                          27.6942      0.438     63.176      0.000      26.835      28.553
no_of_weekend_nights                    -2.3475      0.206    -11.372      0.000      -2.752      -1.943
no_of_week_nights                       -0.2406      0.129     -1.863      0.062      -0.494       0.013
required_car_parking_space               8.8787      1.012      8.773      0.000       6.895      10.862
lead_time                               -0.0525      0.002    -24.103      0.000      -0.057      -0.048
arrival_year                            18.4382      0.502     36.724      0.000      17.454      19.422
arrival_month                            1.5344      0.063     24.297      0.000       1.411       1.658
arrival_date                             0.0186      0.020      0.923      0.356      -0.021       0.058
repeated_guest                         -27.3924      1.368    -20.027      0.000     -30.073     -24.712
no_of_previous_cancellations             1.2455      0.541      2.304      0.021       0.186       2.305
no_of_previous_bookings_not_canceled    -0.6694      0.125     -5.371      0.000      -0.914      -0.425
no_of_special_requests                   2.3632      0.235     10.046      0.000       1.902       2.824
==============================================================================
Omnibus:                     2251.483   Durbin-Watson:                   1.996
Prob(Omnibus):                  0.000   Jarque-Bera (JB):            13283.815
Skew:                           0.026   Prob(JB):                         0.00
Kurtosis:                       6.314   Cond. No.                     1.17e+07
==============================================================================

Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
[2] The condition number is large, 1.17e+07. This might indicate that there are
strong multicollinearity or other numerical problems.


import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LogisticRegression
from sklearn.metrics import accuracy_score, classification_report, confusion_matrix, roc_curve, roc_auc_score, mean_squared_error, mean_absolute_error, r2_score
import statsmodels.api as sm
import numpy as np
import matplotlib.pyplot as plt

# Assuming df is your preprocessed DataFrame

# Ensure target variable is of numeric type
df['booking_status'] = df['booking_status'].astype('category').cat.codes

# Define target and features
target = 'booking_status'
features = ['no_of_adults', 'no_of_children', 'no_of_weekend_nights', 'no_of_week_nights',
            'required_car_parking_space', 'lead_time', 'arrival_year', 'arrival_month',
            'repeated_guest', 'no_of_previous_cancellations',
            'no_of_previous_bookings_not_canceled', 'no_of_special_requests']

# Splitting the data into features and target
X = df[features]
y = df[target]

# Splitting the data into training and testing sets
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42, stratify=y)

# Add constant to the model (intercept)
X_train_final = sm.add_constant(X_train)
X_test_final = sm.add_constant(X_test)

# Ensure all data is numeric
X_train_final = X_train_final.apply(pd.to_numeric)
X_test_final = X_test_final.apply(pd.to_numeric)
y_train = y_train.apply(pd.to_numeric)
y_test = y_test.apply(pd.to_numeric)

# Training the Logistic Regression model using statsmodels
logit_model_final = sm.Logit(y_train, X_train_final).fit()
print(logit_model_final.summary())

# Function to compute model performance metrics
def model_performance_classification(model, predictors, target):
    pred_prob = model.predict(predictors)
    pred = (pred_prob > 0.5).astype(int)
    accuracy = accuracy_score(target, pred)
    roc_auc = roc_auc_score(target, pred_prob)
    mse = mean_squared_error(target, pred)
    mae = mean_absolute_error(target, pred)
    r2 = r2_score(target, pred_prob)
    return pd.DataFrame({
        "Accuracy": [accuracy],
        "ROC-AUC": [roc_auc],
        "MSE": [mse],
        "MAE": [mae],
        "R-squared": [r2]
    })

Optimization terminated successfully.
         Current function value: 0.478136
         Iterations 12
                           Logit Regression Results                           
==============================================================================
Dep. Variable:         booking_status   No. Observations:                29020
Model:                          Logit   Df Residuals:                    29007
Method:                           MLE   Df Model:                           12
Date:                Thu, 11 Jul 2024   Pseudo R-squ.:                  0.2440
Time:                        22:45:15   Log-Likelihood:                -13876.
converged:                       True   LL-Null:                       -18355.
Covariance Type:            nonrobust   LLR p-value:                     0.000
========================================================================================================
                                           coef    std err          z      P>|z|      [0.025      0.975]
--------------------------------------------------------------------------------------------------------
const                                 2419.1975     98.980     24.441      0.000    2225.201    2613.194
no_of_adults                            -0.4394      0.030    -14.623      0.000      -0.498      -0.381
no_of_children                          -0.5134      0.036    -14.150      0.000      -0.585      -0.442
no_of_weekend_nights                    -0.1255      0.017     -7.366      0.000      -0.159      -0.092
no_of_week_nights                       -0.0346      0.011     -3.285      0.001      -0.055      -0.014
required_car_parking_space               1.1290      0.126      8.954      0.000       0.882       1.376
lead_time                               -0.0110      0.000    -56.330      0.000      -0.011      -0.011
arrival_year                            -1.1978      0.049    -24.421      0.000      -1.294      -1.102
arrival_month                           -0.0008      0.005     -0.148      0.882      -0.011       0.010
repeated_guest                           2.4000      0.411      5.843      0.000       1.595       3.205
no_of_previous_cancellations            -0.2293      0.074     -3.114      0.002      -0.374      -0.085
no_of_previous_bookings_not_canceled     0.1161      0.092      1.255      0.209      -0.065       0.297
no_of_special_requests                   1.0426      0.024     42.866      0.000       0.995       1.090
========================================================================================================


import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler, OneHotEncoder
from sklearn.compose import ColumnTransformer
from sklearn.pipeline import Pipeline
from sklearn.impute import SimpleImputer


# Define target and features
target = 'booking_status'  # Assuming 'booking_status' is the target variable for classification
features = ['no_of_adults', 'no_of_children', 'no_of_weekend_nights', 'no_of_week_nights',
            'required_car_parking_space', 'lead_time', 'arrival_year', 'arrival_month',
            'repeated_guest', 'no_of_previous_cancellations',
            'no_of_previous_bookings_not_canceled', 'no_of_special_requests',
            'avg_price_per_room']

# Split data into features and target
X = df[features]
y = df[target]

# Split data into training and testing sets
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

# Data preprocessing pipeline
numeric_features = ['no_of_adults', 'no_of_children', 'no_of_weekend_nights', 'no_of_week_nights',
                    'required_car_parking_space', 'lead_time', 'arrival_year', 'arrival_month',
                    'repeated_guest', 'no_of_previous_cancellations',
                    'no_of_previous_bookings_not_canceled', 'no_of_special_requests',
                    'avg_price_per_room']

numeric_transformer = Pipeline(steps=[
    ('imputer', SimpleImputer(strategy='mean')),
    ('scaler', StandardScaler())
])

preprocessor = ColumnTransformer(
    transformers=[
        ('num', numeric_transformer, numeric_features)
    ])

# Prepare the pipeline for logistic regression
from sklearn.linear_model import LogisticRegression

logreg_pipeline = Pipeline(steps=[('preprocessor', preprocessor),
                                  ('classifier', LogisticRegression(random_state=42, max_iter=1000))])
# Fit the model
logreg_pipeline.fit(X_train, y_train)

Pipeline(steps=[('preprocessor',
                 ColumnTransformer(transformers=[('num',
                                                  Pipeline(steps=[('imputer',
                                                                   SimpleImputer()),
                                                                  ('scaler',
                                                                   StandardScaler())]),
                                                  ['no_of_adults',
                                                   'no_of_children',
                                                   'no_of_weekend_nights',
                                                   'no_of_week_nights',
                                                   'required_car_parking_space',
                                                   'lead_time', 'arrival_year',
                                                   'arrival_month',
                                                   'repeated_guest',
                                                   'no_of_previous_cancellations',
                                                   'no_of_previous_bookings_not_canceled',
                                                   'no_of_special_requests',
                                                   'avg_price_per_room'])])),
                ('classifier',
                 LogisticRegression(max_iter=1000, random_state=42))])

Pipeline(steps=[('preprocessor',
                 ColumnTransformer(transformers=[('num',
                                                  Pipeline(steps=[('imputer',
                                                                   SimpleImputer()),
                                                                  ('scaler',
                                                                   StandardScaler())]),
                                                  ['no_of_adults',
                                                   'no_of_children',
                                                   'no_of_weekend_nights',
                                                   'no_of_week_nights',
                                                   'required_car_parking_space',
                                                   'lead_time', 'arrival_year',
                                                   'arrival_month',
                                                   'repeated_guest',
                                                   'no_of_previous_cancellations',
                                                   'no_of_previous_bookings_not_canceled',
                                                   'no_of_special_requests',
                                                   'avg_price_per_room'])])),
                ('classifier',
                 LogisticRegression(max_iter=1000, random_state=42))])

ColumnTransformer(transformers=[('num',
                                 Pipeline(steps=[('imputer', SimpleImputer()),
                                                 ('scaler', StandardScaler())]),
                                 ['no_of_adults', 'no_of_children',
                                  'no_of_weekend_nights', 'no_of_week_nights',
                                  'required_car_parking_space', 'lead_time',
                                  'arrival_year', 'arrival_month',
                                  'repeated_guest',
                                  'no_of_previous_cancellations',
                                  'no_of_previous_bookings_not_canceled',
                                  'no_of_special_requests',
                                  'avg_price_per_room'])])

['no_of_adults', 'no_of_children', 'no_of_weekend_nights', 'no_of_week_nights', 'required_car_parking_space', 'lead_time', 'arrival_year', 'arrival_month', 'repeated_guest', 'no_of_previous_cancellations', 'no_of_previous_bookings_not_canceled', 'no_of_special_requests', 'avg_price_per_room']

SimpleImputer()

StandardScaler()

LogisticRegression(max_iter=1000, random_state=42)


from sklearn.metrics import classification_report, confusion_matrix, roc_auc_score, roc_curve, accuracy_score

# Predict on the test set
y_pred = logreg_pipeline.predict(X_test)
y_pred_prob = logreg_pipeline.predict_proba(X_test)[:, 1]

# Evaluate the model
print("Accuracy:", accuracy_score(y_test, y_pred))
print("Classification Report:\n", classification_report(y_test, y_pred))
print("Confusion Matrix:\n", confusion_matrix(y_test, y_pred))
print("ROC-AUC Score:", roc_auc_score(y_test, y_pred_prob))

# Plot ROC Curve
import matplotlib.pyplot as plt

fpr, tpr, _ = roc_curve(y_test, y_pred_prob)
plt.figure(figsize=(8, 6))
plt.plot(fpr, tpr, label='Logistic Regression (AUC = {:.2f})'.format(roc_auc_score(y_test, y_pred_prob)))
plt.plot([0, 1], [0, 1], linestyle='--')
plt.xlabel('False Positive Rate')
plt.ylabel('True Positive Rate')
plt.title('ROC Curve')
plt.legend()
plt.show()

Accuracy: 0.7875947622329428
Classification Report:
               precision    recall  f1-score   support

           0       0.74      0.56      0.64      2416
           1       0.80      0.90      0.85      4839

    accuracy                           0.79      7255
   macro avg       0.77      0.73      0.74      7255
weighted avg       0.78      0.79      0.78      7255

Confusion Matrix:
 [[1361 1055]
 [ 486 4353]]
ROC-AUC Score: 0.8377523217812228


print("Columns in X_test:")
print(X_test.columns)

Columns in X_test:
Index(['no_of_adults', 'no_of_children', 'no_of_weekend_nights',
       'no_of_week_nights', 'required_car_parking_space', 'lead_time',
       'arrival_year', 'arrival_month', 'repeated_guest',
       'no_of_previous_cancellations', 'no_of_previous_bookings_not_canceled',
       'no_of_special_requests', 'avg_price_per_room'],
      dtype='object')


import numpy as np
import matplotlib.pyplot as plt
from sklearn.metrics import roc_auc_score, roc_curve, accuracy_score, confusion_matrix
import seaborn as sns

# Print the columns in X_test to identify the correct feature names
print("Columns in X_test:")
print(X_test.columns)

# Assuming you have identified features to drop (replace these with actual feature names if different)
features_to_drop = ['no_of_weekend_nights', 'no_of_week_nights',
                    'required_car_parking_space', 'lead_time',
                    'arrival_year', 'arrival_month', 'repeated_guest',
                    'no_of_previous_cancellations', 'no_of_special_requests']

# Drop the specified features from the test set
X_test1 = X_test.drop(features_to_drop, axis=1)

# Ensure you drop the same features from X_train
X_train1 = X_train.drop(features_to_drop, axis=1)

# Train the Logistic Regression model on the new training set
logit_model_final = sm.Logit(y_train, sm.add_constant(X_train1)).fit()
print(logit_model_final.summary())

# Compute ROC-AUC for the training set
logit_roc_auc_train = roc_auc_score(y_train, logit_model_final.predict(sm.add_constant(X_train1)))
fpr, tpr, thresholds = roc_curve(y_train, logit_model_final.predict(sm.add_constant(X_train1)))

plt.figure(figsize=(7, 5))
plt.plot(fpr, tpr, label="Logistic Regression (area = %0.2f)" % logit_roc_auc_train)
plt.plot([0, 1], [0, 1], "r--")
plt.xlim([0.0, 1.0])
plt.ylim([0.0, 1.05])
plt.xlabel("False Positive Rate")
plt.ylabel("True Positive Rate")
plt.title("Receiver Operating Characteristic - Training Data")
plt.legend(loc="lower right")
plt.show()

# Predict on the modified test set
pred_test = logit_model_final.predict(sm.add_constant(X_test1)) > 0.5
pred_test = np.round(pred_test)

# Evaluate accuracy
print("Accuracy on training set: ", accuracy_score(y_train, logit_model_final.predict(sm.add_constant(X_train1)) > 0.5))
print("Accuracy on test set: ", accuracy_score(y_test, pred_test))

# Plotting the confusion matrix for the test set
cm_test = confusion_matrix(y_test, pred_test)
plt.figure(figsize=(7, 5))
sns.heatmap(cm_test, annot=True, fmt="g")
plt.xlabel("Predicted Values")
plt.ylabel("Actual Values")
plt.title("Confusion Matrix - Test Data")
plt.show()

Columns in X_test:
Index(['no_of_adults', 'no_of_children', 'no_of_weekend_nights',
       'no_of_week_nights', 'required_car_parking_space', 'lead_time',
       'arrival_year', 'arrival_month', 'repeated_guest',
       'no_of_previous_cancellations', 'no_of_previous_bookings_not_canceled',
       'no_of_special_requests', 'avg_price_per_room'],
      dtype='object')
Optimization terminated successfully.
         Current function value: 0.615315
         Iterations 11
                           Logit Regression Results                           
==============================================================================
Dep. Variable:         booking_status   No. Observations:                29020
Model:                          Logit   Df Residuals:                    29015
Method:                           MLE   Df Model:                            4
Date:                Thu, 11 Jul 2024   Pseudo R-squ.:                 0.02566
Time:                        23:06:11   Log-Likelihood:                -17856.
converged:                       True   LL-Null:                       -18327.
Covariance Type:            nonrobust   LLR p-value:                2.598e-202
========================================================================================================
                                           coef    std err          z      P>|z|      [0.025      0.975]
--------------------------------------------------------------------------------------------------------
const                                    1.8116      0.057     32.001      0.000       1.701       1.923
no_of_adults                            -0.1662      0.026     -6.313      0.000      -0.218      -0.115
no_of_children                           0.0877      0.033      2.619      0.009       0.022       0.153
no_of_previous_bookings_not_canceled     1.1931      0.168      7.110      0.000       0.864       1.522
avg_price_per_room                      -0.0077      0.000    -18.478      0.000      -0.009      -0.007
========================================================================================================

Accuracy on training set:  0.6702274293590628
Accuracy on test set:  0.6638180565127498


import pandas as pd
import numpy as np
from sklearn.tree import DecisionTreeClassifier, plot_tree, export_text
from sklearn.model_selection import train_test_split
from sklearn import metrics
import matplotlib.pyplot as plt
import seaborn as sns

# Assuming df is your preprocessed DataFrame and booking_status is already numeric
# Define target and features
target = 'booking_status'
features = ['no_of_adults', 'no_of_children', 'required_car_parking_space', 'lead_time',
            'arrival_month', 'repeated_guest', 'no_of_previous_cancellations',
            'no_of_previous_bookings_not_canceled', 'no_of_special_requests']

# Prepare the data
tree_data = df[features + [target]].astype(float)

# Drop irrelevant features
tree_data = tree_data.drop(['arrival_date', 'arrival_year', 'no_of_week_nights', 'no_of_weekend_nights'], axis=1, errors='ignore')

# Split the data into features and target
X = tree_data.drop(target, axis=1)
y = tree_data[target]

# Split the data into training and testing sets
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.30, random_state=1)

# Initialize and fit the Decision Tree model with pre-pruning
dTree = DecisionTreeClassifier(criterion='gini', random_state=1, max_depth=5, min_samples_split=20, min_samples_leaf=5)
dTree.fit(X_train, y_train)

# Print accuracy
print("Accuracy on training set: ", dTree.score(X_train, y_train))
print("Accuracy on test set: ", dTree.score(X_test, y_test))

# Function to make confusion matrix
def make_confusion_matrix(model, y_actual, X_test, labels=[1, 0]):
    y_predict = model.predict(X_test)
    cm = metrics.confusion_matrix(y_actual, y_predict, labels=labels)
    df_cm = pd.DataFrame(cm, index=["Actual - No", "Actual - Yes"], columns=['Predicted - No', 'Predicted - Yes'])
    group_counts = ["{0:0.0f}".format(value) for value in cm.flatten()]
    group_percentages = ["{0:.2%}".format(value) for value in cm.flatten() / np.sum(cm)]
    labels = [f"{v1}\n{v2}" for v1, v2 in zip(group_counts, group_percentages)]
    labels = np.asarray(labels).reshape(2, 2)
    plt.figure(figsize=(10, 7))
    sns.heatmap(df_cm, annot=labels, fmt='', cmap='Blues')
    plt.ylabel('True label')
    plt.xlabel('Predicted label')
    plt.title('Confusion Matrix')
    plt.show()

# Function to calculate recall score
def get_recall_score(model, X_train, y_train, X_test, y_test):
    pred_train = model.predict(X_train)
    pred_test = model.predict(X_test)
    print("Recall on training set: ", metrics.recall_score(y_train, pred_train))
    print("Recall on test set: ", metrics.recall_score(y_test, pred_test))

# Generate confusion matrix for test set
make_confusion_matrix(dTree, y_test, X_test)

# Calculate recall score
get_recall_score(dTree, X_train, y_train, X_test, y_test)

# Plotting Feature Importance
feature_names = list(X_train.columns)
importances = dTree.feature_importances_
indices = np.argsort(importances)

plt.figure(figsize=(10, 6))
plt.title('Feature Importances')
plt.barh(range(len(indices)), importances[indices], align='center')
plt.yticks(range(len(indices)), [feature_names[i] for i in indices])
plt.xlabel('Relative Importance')
plt.show()

# Plotting Decision Tree
plt.figure(figsize=(20, 10))
out = plot_tree(
    dTree,
    feature_names=feature_names,
    filled=True,
    fontsize=9,
    node_ids=False,
    class_names=['Not_Canceled', 'Canceled']
)

# Add arrows to the decision tree split if they are missing
for o in out:
    arrow = o.arrow_patch
    if arrow is not None:
        arrow.set_edgecolor("black")
        arrow.set_linewidth(1)

plt.show()

# Text report showing the rules of a decision tree
tree_rules = export_text(dTree, feature_names=feature_names, show_weights=True)
print(tree_rules)

Accuracy on training set:  0.7756379962192816
Accuracy on test set:  0.7782780483322613

Recall on training set:  0.9325855892888602
Recall on test set:  0.9330254041570438

|--- lead_time <= 151.50
|   |--- no_of_special_requests <= 0.50
|   |   |--- lead_time <= 14.50
|   |   |   |--- arrival_month <= 8.50
|   |   |   |   |--- arrival_month <= 1.50
|   |   |   |   |   |--- weights: [0.00, 211.00] class: 1.0
|   |   |   |   |--- arrival_month >  1.50
|   |   |   |   |   |--- weights: [400.00, 1202.00] class: 1.0
|   |   |   |--- arrival_month >  8.50
|   |   |   |   |--- arrival_month <= 11.50
|   |   |   |   |   |--- weights: [87.00, 844.00] class: 1.0
|   |   |   |   |--- arrival_month >  11.50
|   |   |   |   |   |--- weights: [0.00, 302.00] class: 1.0
|   |   |--- lead_time >  14.50
|   |   |   |--- arrival_month <= 8.50
|   |   |   |   |--- arrival_month <= 1.50
|   |   |   |   |   |--- weights: [16.00, 239.00] class: 1.0
|   |   |   |   |--- arrival_month >  1.50
|   |   |   |   |   |--- weights: [2207.00, 2251.00] class: 1.0
|   |   |   |--- arrival_month >  8.50
|   |   |   |   |--- lead_time <= 92.50
|   |   |   |   |   |--- weights: [505.00, 1786.00] class: 1.0
|   |   |   |   |--- lead_time >  92.50
|   |   |   |   |   |--- weights: [334.00, 283.00] class: 0.0
|   |--- no_of_special_requests >  0.50
|   |   |--- no_of_special_requests <= 1.50
|   |   |   |--- lead_time <= 8.50
|   |   |   |   |--- lead_time <= 4.50
|   |   |   |   |   |--- weights: [29.00, 906.00] class: 1.0
|   |   |   |   |--- lead_time >  4.50
|   |   |   |   |   |--- weights: [44.00, 438.00] class: 1.0
|   |   |   |--- lead_time >  8.50
|   |   |   |   |--- required_car_parking_space <= 0.50
|   |   |   |   |   |--- weights: [981.00, 4084.00] class: 1.0
|   |   |   |   |--- required_car_parking_space >  0.50
|   |   |   |   |   |--- weights: [1.00, 196.00] class: 1.0
|   |   |--- no_of_special_requests >  1.50
|   |   |   |--- lead_time <= 90.50
|   |   |   |   |--- lead_time <= 8.50
|   |   |   |   |   |--- weights: [2.00, 596.00] class: 1.0
|   |   |   |   |--- lead_time >  8.50
|   |   |   |   |   |--- weights: [36.00, 1842.00] class: 1.0
|   |   |   |--- lead_time >  90.50
|   |   |   |   |--- no_of_special_requests <= 2.50
|   |   |   |   |   |--- weights: [107.00, 391.00] class: 1.0
|   |   |   |   |--- no_of_special_requests >  2.50
|   |   |   |   |   |--- weights: [0.00, 90.00] class: 1.0
|--- lead_time >  151.50
|   |--- no_of_adults <= 1.50
|   |   |--- lead_time <= 189.00
|   |   |   |--- lead_time <= 165.00
|   |   |   |   |--- lead_time <= 163.50
|   |   |   |   |   |--- weights: [61.00, 11.00] class: 0.0
|   |   |   |   |--- lead_time >  163.50
|   |   |   |   |   |--- weights: [9.00, 64.00] class: 1.0
|   |   |   |--- lead_time >  165.00
|   |   |   |   |--- no_of_special_requests <= 0.50
|   |   |   |   |   |--- weights: [198.00, 10.00] class: 0.0
|   |   |   |   |--- no_of_special_requests >  0.50
|   |   |   |   |   |--- weights: [18.00, 13.00] class: 0.0
|   |   |--- lead_time >  189.00
|   |   |   |--- lead_time <= 255.50
|   |   |   |   |--- lead_time <= 192.50
|   |   |   |   |   |--- weights: [7.00, 61.00] class: 1.0
|   |   |   |   |--- lead_time >  192.50
|   |   |   |   |   |--- weights: [116.00, 77.00] class: 0.0
|   |   |   |--- lead_time >  255.50
|   |   |   |   |--- lead_time <= 341.00
|   |   |   |   |   |--- weights: [15.00, 154.00] class: 1.0
|   |   |   |   |--- lead_time >  341.00
|   |   |   |   |   |--- weights: [30.00, 19.00] class: 0.0
|   |--- no_of_adults >  1.50
|   |   |--- arrival_month <= 11.50
|   |   |   |--- no_of_special_requests <= 0.50
|   |   |   |   |--- lead_time <= 271.00
|   |   |   |   |   |--- weights: [1255.00, 202.00] class: 0.0
|   |   |   |   |--- lead_time >  271.00
|   |   |   |   |   |--- weights: [710.00, 52.00] class: 0.0
|   |   |   |--- no_of_special_requests >  0.50
|   |   |   |   |--- no_of_special_requests <= 2.50
|   |   |   |   |   |--- weights: [1078.00, 480.00] class: 0.0
|   |   |   |   |--- no_of_special_requests >  2.50
|   |   |   |   |   |--- weights: [0.00, 57.00] class: 1.0
|   |   |--- arrival_month >  11.50
|   |   |   |--- lead_time <= 207.50
|   |   |   |   |--- lead_time <= 165.00
|   |   |   |   |   |--- weights: [2.00, 37.00] class: 1.0
|   |   |   |   |--- lead_time >  165.00
|   |   |   |   |   |--- weights: [16.00, 39.00] class: 1.0
|   |   |   |--- lead_time >  207.50
|   |   |   |   |--- lead_time <= 324.00
|   |   |   |   |   |--- weights: [85.00, 91.00] class: 1.0
|   |   |   |   |--- lead_time >  324.00
|   |   |   |   |   |--- weights: [14.00, 1.00] class: 0.0


dTree1 = DecisionTreeClassifier(criterion = 'gini',max_depth=3,random_state=1)
dTree1.fit(X_train, y_train)

DecisionTreeClassifier(max_depth=3, random_state=1)

DecisionTreeClassifier(max_depth=3, random_state=1)


import pandas as pd
import numpy as np
from sklearn.tree import DecisionTreeClassifier, plot_tree
from sklearn.model_selection import train_test_split
from sklearn import metrics
import matplotlib.pyplot as plt
import seaborn as sns

# Assuming df is your preprocessed DataFrame and booking_status is already numeric

# Define target and features
target = 'booking_status'
features = ['no_of_adults', 'no_of_children', 'no_of_weekend_nights', 'no_of_week_nights',
            'required_car_parking_space', 'lead_time', 'arrival_year', 'arrival_month',
            'repeated_guest', 'no_of_previous_cancellations',
            'no_of_previous_bookings_not_canceled', 'no_of_special_requests']

# Prepare the data
tree_data = df[features + [target]].astype(float)

# Drop irrelevant features
tree_data = tree_data.drop(['arrival_date', 'arrival_year', 'no_of_week_nights', 'no_of_weekend_nights'], axis=1, errors='ignore')

# Split the data into features and target
X = tree_data.drop(target, axis=1)
y = tree_data[target]

# Split the data into training and testing sets
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.30, random_state=1)

# Initialize and fit the Decision Tree model with max depth of 3
dTree1 = DecisionTreeClassifier(criterion='gini', max_depth=3, random_state=1)
dTree1.fit(X_train, y_train)

# Function to make confusion matrix
def make_confusion_matrix_sklearn(model, X, y):
    y_pred = model.predict(X)
    cm = metrics.confusion_matrix(y, y_pred)
    sns.heatmap(cm, annot=True, fmt='g')
    plt.xlabel("Predicted Values")
    plt.ylabel("Actual Values")
    plt.title("Confusion Matrix")
    plt.show()

# Generate confusion matrix for the test set
make_confusion_matrix_sklearn(dTree1, X_test, y_test)

# Print accuracy
print("Accuracy on training set: ", dTree1.score(X_train, y_train))
print("Accuracy on test set: ", dTree1.score(X_test, y_test))

# Plotting Feature Importances
feature_names = list(X_train.columns)
importances = dTree1.feature_importances_
indices = np.argsort(importances)

plt.figure(figsize=(12, 12))
plt.title("Feature Importances")
plt.barh(range(len(indices)), importances[indices], color="violet", align="center")
plt.yticks(range(len(indices)), [feature_names[i] for i in indices])
plt.xlabel("Relative Importance")
plt.show()

# Plotting Decision Tree
plt.figure(figsize=(20, 10))
out = plot_tree(
    dTree1,
    feature_names=feature_names,
    filled=True,
    fontsize=9,
    node_ids=False,
    class_names=['Not_Canceled', 'Canceled']
)

# Add arrows to the decision tree split if they are missing
for o in out:
    arrow = o.arrow_patch
    if arrow is not None:
        arrow.set_edgecolor("black")
        arrow.set_linewidth(1)

plt.show()

Accuracy on training set:  0.7667375551354757
Accuracy on test set:  0.7701920426353027


from sklearn.tree import DecisionTreeClassifier

# Fit the initial decision tree
clf = DecisionTreeClassifier(random_state=1)
clf.fit(X_train, y_train)

# Prune the tree using cost complexity pruning
path = clf.cost_complexity_pruning_path(X_train, y_train)
ccp_alphas, impurities = path.ccp_alphas, path.impurities

# Plot the total impurity vs effective alpha
fig, ax = plt.subplots(figsize=(10, 5))
ax.plot(ccp_alphas[:-1], impurities[:-1], marker='o', drawstyle="steps-post")
ax.set_xlabel("Effective Alpha")
ax.set_ylabel("Total Impurity of Leaves")
ax.set_title("Total Impurity vs Effective Alpha for Training Set")
plt.show()


import numpy as np
import matplotlib.pyplot as plt
from sklearn.tree import DecisionTreeClassifier

# Fit the initial decision tree
clf = DecisionTreeClassifier(random_state=1)
clf.fit(X_train, y_train)

# Prune the tree using cost complexity pruning
path = clf.cost_complexity_pruning_path(X_train, y_train)
ccp_alphas, impurities = path.ccp_alphas, path.impurities

# Filter out non-valid (negative) ccp_alpha values
valid_ccp_alphas = [alpha for alpha in ccp_alphas if alpha >= 0]

# Select a subset of ccp_alphas to speed up the process
subset_ccp_alphas = valid_ccp_alphas[::10]  # Select every 10th value for example

# Decision Tree classifier for every valid alpha
clfs = []
for ccp_alpha in subset_ccp_alphas:
    clf = DecisionTreeClassifier(random_state=1, ccp_alpha=ccp_alpha)
    clf.fit(X_train, y_train)
    clfs.append(clf)

print("Number of nodes in the last tree is: {} with ccp_alpha: {}".format(
    clfs[-1].tree_.node_count, subset_ccp_alphas[-1]))

# Remove the last element which is the trivial tree with one node
clfs = clfs[:-1]
subset_ccp_alphas = subset_ccp_alphas[:-1]

node_counts = [clf.tree_.node_count for clf in clfs]
depth = [clf.tree_.max_depth for clf in clfs]

# Plotting
fig, ax = plt.subplots(2, 1, figsize=(10, 7))

ax[0].plot(subset_ccp_alphas, node_counts, marker='o', drawstyle="steps-post")
ax[0].set_xlabel("Alpha")
ax[0].set_ylabel("Number of Nodes")
ax[0].set_title("Number of Nodes vs Alpha")

ax[1].plot(subset_ccp_alphas, depth, marker='o', drawstyle="steps-post")
ax[1].set_xlabel("Alpha")
ax[1].set_ylabel("Depth of Tree")
ax[1].set_title("Depth vs Alpha")

fig.tight_layout()
plt.show()

Number of nodes in the last tree is: 5 with ccp_alpha: 0.010030306239306092


import numpy as np
import matplotlib.pyplot as plt
from sklearn.tree import DecisionTreeClassifier

# Fit the initial decision tree
clf = DecisionTreeClassifier(random_state=1)
clf.fit(X_train, y_train)

# Prune the tree using cost complexity pruning
path = clf.cost_complexity_pruning_path(X_train, y_train)
ccp_alphas, impurities = path.ccp_alphas, path.impurities

# Filter out non-valid (negative) ccp_alpha values
valid_ccp_alphas = [alpha for alpha in ccp_alphas if alpha >= 0]

# Select a subset of ccp_alphas to speed up the process
subset_ccp_alphas = valid_ccp_alphas[::10]  # Select every 10th value for example

# Decision Tree classifier for every valid alpha
clfs = []
for ccp_alpha in subset_ccp_alphas:
    clf = DecisionTreeClassifier(random_state=1, ccp_alpha=ccp_alpha)
    clf.fit(X_train, y_train)
    clfs.append(clf)

print("Number of nodes in the last tree is: {} with ccp_alpha: {}".format(
    clfs[-1].tree_.node_count, subset_ccp_alphas[-1]))

# Remove the last element which is the trivial tree with one node
clfs = clfs[:-1]
subset_ccp_alphas = subset_ccp_alphas[:-1]

train_scores = [clf.score(X_train, y_train) for clf in clfs]
test_scores = [clf.score(X_test, y_test) for clf in clfs]

# Plotting accuracy vs alpha
fig, ax = plt.subplots(figsize=(10, 5))
ax.set_xlabel("Alpha")
ax.set_ylabel("Accuracy")
ax.set_title("Accuracy vs Alpha for Training and Testing Sets")
ax.plot(subset_ccp_alphas, train_scores, marker='o', label="Train", drawstyle="steps-post")
ax.plot(subset_ccp_alphas, test_scores, marker='o', label="Test", drawstyle="steps-post")
ax.legend()
plt.show()

Number of nodes in the last tree is: 5 with ccp_alpha: 0.010030306239306092


# Selecting the best model based on test scores
index_best_model = np.argmax(test_scores)
best_model = clfs[index_best_model]
print("Best Decision Tree Model:\n", best_model)
print('Training accuracy of best model: ', best_model.score(X_train, y_train))
print('Test accuracy of best model: ', best_model.score(X_test, y_test))

# Recall for training set
recall_train = []
for clf in clfs:
    pred_train3 = clf.predict(X_train)
    values_train = metrics.recall_score(y_train, pred_train3)
    recall_train.append(values_train)

# Recall for testing set
recall_test = []
for clf in clfs:
    pred_test3 = clf.predict(X_test)
    values_test = metrics.recall_score(y_test, pred_test3)
    recall_test.append(values_test)

# Plotting recall vs alpha
fig, ax = plt.subplots(figsize=(15, 5))
ax.set_xlabel("alpha")
ax.set_ylabel("Recall")
ax.set_title("Recall vs alpha for training and testing sets")
ax.plot(subset_ccp_alphas, recall_train, marker='o', label="train", drawstyle="steps-post")
ax.plot(subset_ccp_alphas, recall_test, marker='o', label="test", drawstyle="steps-post")
ax.legend()
plt.show()

Best Decision Tree Model:
 DecisionTreeClassifier(ccp_alpha=8.751662815935032e-05, random_state=1)
Training accuracy of best model:  0.8482592942659105
Test accuracy of best model:  0.8278967196545071


# Assuming 'recall_test' and 'clfs' have been defined in previous cells

# Creating the model where we get the highest train and test recall
index_best_model = np.argmax(recall_test)
best_model = clfs[index_best_model]
print("Best Decision Tree Model:\n", best_model)

# Evaluating the best model
print('Training accuracy of best model: ', best_model.score(X_train, y_train))
print('Test accuracy of best model: ', best_model.score(X_test, y_test))

# Evaluating model performance on the training set
train_performance = model_performance_classification_sklearn(best_model, X_train, y_train)
print("Training Performance\n", train_performance)

# Evaluating model performance on the test set
test_performance = model_performance_classification_sklearn(best_model, X_test, y_test)
print("Test Performance\n", test_performance)

# Plotting feature importance for the best model
importances = best_model.feature_importances_
indices = np.argsort(importances)
plt.figure(figsize=(12, 12))
plt.title("Feature Importances")
plt.barh(range(len(indices)), importances[indices], color="violet", align="center")
plt.yticks(range(len(indices)), [feature_names[i] for i in indices])
plt.xlabel("Relative Importance")
plt.show()

Best Decision Tree Model:
 DecisionTreeClassifier(ccp_alpha=0.001528941615543386, random_state=1)
Training accuracy of best model:  0.7747321991178324
Test accuracy of best model:  0.7764403197647708
Training Performance
    Accuracy    Recall   ROC-AUC
0  0.774732  0.941335  0.818325
Test Performance
    Accuracy    Recall   ROC-AUC
0   0.77644  0.939954  0.815768


import numpy as np
import matplotlib.pyplot as plt
from sklearn import metrics
import seaborn as sns

# Function to make confusion matrix
def make_confusion_matrix_sklearn(model, X, y):
    y_pred = model.predict(X)
    cm = metrics.confusion_matrix(y, y_pred)
    sns.heatmap(cm, annot=True, fmt='g')
    plt.xlabel("Predicted Values")
    plt.ylabel("Actual Values")
    plt.title("Confusion Matrix")
    plt.show()

# Generate confusion matrix for the best model on the test set
make_confusion_matrix_sklearn(best_model, X_test, y_test)


the_features = X_train.columns

plt.figure(figsize=(17, 15))
plot_tree(
    best_model,
    feature_names=the_features,
    filled=True,
    fontsize=9,
    node_ids=True,
    class_names=True
)
plt.show()


import pandas as pd
from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score, roc_auc_score

# Function to compute performance metrics for a classification model
def compute_performance_metrics(model, X_train, y_train, X_test, y_test):
    metrics_dict = {}

    # Training set predictions
    y_train_pred = model.predict(X_train)
    y_train_prob = model.predict_proba(X_train)[:, 1]

    # Test set predictions
    y_test_pred = model.predict(X_test)
    y_test_prob = model.predict_proba(X_test)[:, 1]

    # Training set performance
    metrics_dict['Train Accuracy'] = accuracy_score(y_train, y_train_pred)
    metrics_dict['Train Precision'] = precision_score(y_train, y_train_pred)
    metrics_dict['Train Recall'] = recall_score(y_train, y_train_pred)
    metrics_dict['Train F1 Score'] = f1_score(y_train, y_train_pred)
    metrics_dict['Train ROC AUC'] = roc_auc_score(y_train, y_train_prob)

    # Test set performance
    metrics_dict['Test Accuracy'] = accuracy_score(y_test, y_test_pred)
    metrics_dict['Test Precision'] = precision_score(y_test, y_test_pred)
    metrics_dict['Test Recall'] = recall_score(y_test, y_test_pred)
    metrics_dict['Test F1 Score'] = f1_score(y_test, y_test_pred)
    metrics_dict['Test ROC AUC'] = roc_auc_score(y_test, y_test_prob)

    return metrics_dict

# Compute performance metrics for the best decision tree model
decision_tree_metrics = compute_performance_metrics(best_model, X_train, y_train, X_test, y_test)

# Print performance metrics
decision_tree_metrics_df = pd.DataFrame.from_dict(decision_tree_metrics, orient='index', columns=['Decision Tree'])
print(decision_tree_metrics_df)

# You can add other models' metrics to this DataFrame for comparison
# For example, adding logistic regression model's metrics
# logistic_regression_metrics = compute_performance_metrics(logistic_regression_model, X_train, y_train, X_test, y_test)
# decision_tree_metrics_df['Logistic Regression'] = pd.Series(logistic_regression_metrics)

# Print comparison table
print(decision_tree_metrics_df)

                 Decision Tree
Train Accuracy        0.774732
Train Precision       0.772493
Train Recall          0.941335
Train F1 Score        0.848597
Train ROC AUC         0.818325
Test Accuracy         0.776440
Test Precision        0.776543
Test Recall           0.939954
Test F1 Score         0.850470
Test ROC AUC          0.815768
                 Decision Tree
Train Accuracy        0.774732
Train Precision       0.772493
Train Recall          0.941335
Train F1 Score        0.848597
Train ROC AUC         0.818325
Test Accuracy         0.776440
Test Precision        0.776543
Test Recall           0.939954
Test F1 Score         0.850470
Test ROC AUC          0.815768


# Summarizing  the performance of the best model
best_model_performance = decision_tree_metrics_df['Decision Tree']

print("\nModel Performance Summary:")
print(f"Train Accuracy: {best_model_performance['Train Accuracy']:.2f}")
print(f"Test Accuracy: {best_model_performance['Test Accuracy']:.2f}")
print(f"Train Precision: {best_model_performance['Train Precision']:.2f}")
print(f"Test Precision: {best_model_performance['Test Precision']:.2f}")
print(f"Train Recall: {best_model_performance['Train Recall']:.2f}")
print(f"Test Recall: {best_model_performance['Test Recall']:.2f}")
print(f"Train F1 Score: {best_model_performance['Train F1 Score']:.2f}")
print(f"Test F1 Score: {best_model_performance['Test F1 Score']:.2f}")
print(f"Train ROC AUC: {best_model_performance['Train ROC AUC']:.2f}")
print(f"Test ROC AUC: {best_model_performance['Test ROC AUC']:.2f}")

# Conclusions
print("\nConclusions:")
print("1. The decision tree model shows high accuracy on both the training and test sets, indicating good generalization.")
print("2. The recall scores are particularly high, suggesting that the model is effective in identifying positive cases.")
print("3. The ROC AUC score is also high, indicating a strong ability to distinguish between the classes.")
print("4. Overall, the decision tree model performs well, but it is important to monitor for potential overfitting.")
print("5. Further tuning of hyperparameters and possibly pruning the tree could help in improving the model performance.")

Model Performance Summary:
Train Accuracy: 0.77
Test Accuracy: 0.78
Train Precision: 0.77
Test Precision: 0.78
Train Recall: 0.94
Test Recall: 0.94
Train F1 Score: 0.85
Test F1 Score: 0.85
Train ROC AUC: 0.82
Test ROC AUC: 0.82

Conclusions:
1. The decision tree model shows high accuracy on both the training and test sets, indicating good generalization.
2. The recall scores are particularly high, suggesting that the model is effective in identifying positive cases.
3. The ROC AUC score is also high, indicating a strong ability to distinguish between the classes.
4. Overall, the decision tree model performs well, but it is important to monitor for potential overfitting.
5. Further tuning of hyperparameters and possibly pruning the tree could help in improving the model performance.

	Booking_ID	no_of_adults	no_of_weekend_nights	no_of_week_nights	type_of_meal_plan	room_type_reserved	lead_time	arrival_year	arrival_month	arrival_date	market_segment_type	avg_price_per_room	no_of_special_requests	booking_status
0	INN00001	2	1	2	Meal Plan 1	Room_Type 1	224	2017	10	2	Offline	65.00	0	Not_Canceled
1	INN00002	2	2	3	Not Selected	Room_Type 1	5	2018	11	6	Online	106.68	1	Not_Canceled
2	INN00003	1	2	1	Meal Plan 1	Room_Type 1	1	2018	2	28	Online	60.00	0	Canceled
3	INN00004	2	0	2	Meal Plan 1	Room_Type 1	211	2018	5	20	Online	100.00	0	Canceled
4	INN00005	2	1	1	Not Selected	Room_Type 1	48	2018	4	11	Online	94.50	0	Canceled

	count	unique	top	freq
Booking_ID	36275	36275	INN00001	1
type_of_meal_plan	36275	4	Meal Plan 1	27835
room_type_reserved	36275	7	Room_Type 1	28130
market_segment_type	36275	5	Online	23214
booking_status	36275	2	Not_Canceled	24390

	Booking_ID	no_of_adults	no_of_weekend_nights	no_of_week_nights	lead_time	arrival_year	arrival_month	arrival_date	...	market_segment_type_3	market_segment_type_4
0	0.000000	0.50	0.142857	0.117647	0.820513	0.0	0.818182	0.033333	...	1.0	0.0
1	0.000028	0.50	0.285714	0.176471	0.018315	1.0	0.909091	0.166667	...	0.0	1.0
2	0.000055	0.25	0.285714	0.058824	0.003663	1.0	0.090909	0.900000	...	0.0	1.0
3	0.000083	0.50	0.000000	0.117647	0.772894	1.0	0.363636	0.633333	...	0.0	1.0
4	0.000110	0.50	0.142857	0.058824	0.175824	1.0	0.272727	0.333333	...	0.0	1.0

INN Hotels Project¶

Description¶

Context¶

Objective¶

Checking for Missing Values¶

Observations¶

Statistical Summary of Numeric variables¶

Observation¶

Statistical Summary of Categorical variables¶

Observation¶

General observations about the hotel bookings:¶

Check for duplicates¶

Exploratory Data Analysis (EDA)¶

General Observations from Histograms, Box Plots, and Bar Plots¶

Bivariate Analysis¶

General Inference¶

What are the busiest months in the hotel?¶

Observations:¶

Which market segment do most of the guests come from?¶

3.Hotel rates are dynamic and change according to demand and customer demographics. What are the differences in room prices in different market segments?¶

Observations¶

What percentage of bookings are canceled?¶

Observations:¶

Repeating guests are the guests who stay in the hotel often and are important to brand equity. What percentage of repeating guests cancel?¶

Many guests have special requirements when booking a hotel room. Do these requirements affect booking cancellation?¶

Data Pre processing¶

Checking for missing values¶

Encoding Categorical Columns¶

Converting categorical data into numeric format using techniques like one-hot encoding or label encoding. ¶

Feature Engineering ¶

Outlier Detection¶

Conclusion¶

Exploratory Data analysis after Manipulation¶

General Observations from EDA¶

Data Distribution:¶

Outliers:¶

Booking Patterns:¶

Special Requests and Parking:¶

Customer Behavior:¶

Lead Time:¶

Correlation Insights:¶

TEST FOR MULTICOLLINEARITY¶

High Multicollinearity:¶

Moderate Multicollinearity:¶

Low Multicollinearity:¶

Removing Multicollinearity¶

Observations¶

**Building a Logistic Regression model¶

Model performance evaluation¶

Observations ¶

Final Model Summary¶

Building a Decision Tree model¶

Do we need to prune the tree?¶

Model Performance Comparison and Conclusions¶

Conclusion¶

Actionable Insights and Recommendations¶